Government funding agencies spend significant amounts of R&D funds through funding programs. While allocating funds among sectors or scientific disciplines, the Decision Maker (DM) wants to maximize the total impact by supporting R&D activities in those sectors with higher scientific, social and economic return. On the other hand, the DM wants to balance the funding budget over sectors or disciplines. In this study, we incorporate the results of "sectoral impact assessments" into the public R&D project portfolio selection (RDPPS) problem. We develop a two-stage model. In the first stage, we make sectoral budget allocation decisions to maximize the total impact of the budget while ensuring a relative balance among sectors. In the second stage, we maximize the total score of supported projects under allocated sectoral budgets. We illustrate the proposed approach on an example problem. We show the value of the proposed approach by comparing our results with alternative policy options.